Faulty Dive log

[Jef Driesen]

On 02-05-14 03:29, Linus Torvalds wrote:
> On Thu, May 1, 2014 at 5:54 PM, Michael Likes <jetskivet at yahoo.com> wrote:
>>
>> Attached is the screen shot of the stacked dives
>
> That is quite impressive.
>
> It certainly *looks* like we imported all those dives as one single
> dive, and then had a really odd depth spike in between dives. In
> addition, those first three dives seem to be each about 700 _minutes_
> long, which is certainly possible with a rebreather, but judging from
> your pressure graphs I'm assuming that's not it. So the sample rate is
> wrong too. The impressive depth spike between dives is presumably then
> the between-the-dive data being interpreted as bogus sample data with
> random depths.
>
> So at a guess, I'd say that libdivecomputer is terminally confuse,
> possible due to some dive computer memory wrap-around issue. Jef?

That's certainly possible.

> Hopefully Jef will make more sense of your libdivecomputer dump.

@Michael: Can you also enable the libdivecomputer logfile? I need some piece of 
data that is not part of the memory dump, but is logged to the logfile. It's 
version string that is used to select the correct memory layout, so it's 
important info for this his kind issue.

Jef